Advancing Alzheimer Research with AI
UC Davis Team Leads New Effort in Neuropathology Analysis for individuals who identify as Hispanic/Latino
Chen-Nee Chuah, professor of electrical and computer engineering, is part of a multidisciplinary team of researchers that has received a $6 million grant from the National Institute on Aging to provide more automated ways to quantify pathologies in the brain of individuals with Alzheimer disease, especially those who identify as Hispanic and Latino.
The large-scale, five-year initiative aims to present a comprehensive description of the pathological manifestations of Alzheimer disease in the brain of individuals of Mexican, Cuban, Puerto Rican and Dominican heritage in order to develop more tailored precision medicine approaches. The effort is led by Chuah's longtime collaborator Dr. Brittany Dugger, leader of the Neuropathology Core at the UC Davis Alzheimer's Disease Research Center, and features additional faculty from UC Davis Health as well as collaborators from Emory University and the Human Computation Institute.
"Most works on dementia have been focused on persons of certain demographics," Dugger said. "It is imperative to study persons from diverse backgrounds, including persons who identify as Hispanic/Latino, to develop cures that benefit all."
Chuah will lead the development of a machine learning, or ML, process for quantifying pathologies of Alzheimer disease in multiple neuroanatomic locations. The ML model will analyze images of brain tissue to assess distributions of amyloid plaques — one of the pathologic hallmarks of Alzheimer disease — in grey and white matter regions of the brain. Under Chuah's guidance, advanced artificial intelligence and ML techniques will play a pivotal role in enhancing the precision and depth of neuropathologic analysis in underrecognized communities.
"There are currently no standard measures of safety for AI/ML models in terms of robustness against bias, generalizability, or explainability," Chuah said. "We attempt to address these gaps in this project by leveraging brain tissue images from diverse cohorts. It is very rewarding for me and students in the laboratory to contribute to the mission of advancing human health for all with computational methods."
This work will leverage findings from a previous UC Noyce Initiative project led by Chuah in collaboration with Dugger to develop an accessible, interpretable, and efficient deep learning framework for neuropathology and neuroradiology image analysis.
